Open Chemistry 2021; 19: 347–357

Research Article

Belgin Sever, Mehlika Dilek Altıntop*, Yeliz Demir, Cüneyt Türkeş, Kaan Özbaş, Gülşen Akalın Çiftçi, Şükrü Beydemir*, Ahmet Özdemir A new series of 2,4-thiazolidinediones endowed with potent inhibitory activity

https://doi.org/10.1515/chem-2021-0032 received December 2, 2020; accepted February 9, 2021 1 Introduction

Abstract: In an effort to identify potent aldose reductase Type 2 diabetes (T2D) is a chronic life-threatening disease (AR) inhibitors, 5-(arylidene)thiazolidine-2,4-diones (1–8), characterized by abnormally high blood glucose levels which were prepared by the solvent-free reaction of 2,4- resulting from impaired response of target tissues to thiazolidinedione with aromatic aldehydes in the presence insulin (insulin resistance) and/or progressively reduced in vitro of urea, were examined for their AR inhibitory function of pancreatic β cells. The global burden of T2D is -( - - - - activities and cytotoxicity. 5 2 Hydroxy 3 methylbenzyli increasing considerably, and therefore there is an urgent ) - - (3) dene thiazolidine 2,4 dione was the most potent AR need to develop safe and potent antidiabetic agents [1–5]. inhibitor in this series, exerting uncompetitive inhibition Polyol pathway is a two-step metabolic pathway K ± with a i value of 0.445 0.013 µM. The IC50 value of in which glucose is reduced to sorbitol, which is then 3 fi - compound for L929 mouse broblast cells was deter converted to fructose. The abnormally activated polyol mined as 8.9 ± 0.66 µM, pointing out its safety as an AR pathway has been reported to participate in the patho- inhibitor. Molecular docking studies suggested that com- genesis of T2D complications [5–11]. pound 3 exhibited good affinity to the of AR Aldose reductase (AR) catalyzes the NADPH-depen- (PDB ID: 4JIR). Based upon in silico absorption, distribu- dent reduction of glucose to sorbitol as the first and rate- tion, metabolism, and excretion data, the compound is limiting of the polyol pathway. Under euglycemic predicted to have favorable pharmacokinetic features. conditions, the reduction of glucose is a minor function Taking into account the in silico and in vitro data, com- of AR owing to the relatively low affinity (high K ) of AR pound 3 stands out as a potential orally bioavailable AR m for this . However, under hyperglycemic condi- inhibitor for the management of diabetic complications as tions, excess intracellular glucose leads to an increase in well as nondiabetic diseases. the enzymatic conversion of glucose to sorbitol, NADPH- Keywords: aldose reductase, thiazolidinedione, cytotoxi- consuming reaction in tissues possessing insulin-inde- city, molecular docking pendent glucose transport. Sorbitol does not diffuse readily through cell membranes due to its strong hydro- philic feature and accumulates in cells causing osmotic  stress and cellular damage, particularly in lenses. Further- * Corresponding author: Mehlika Dilek Altıntop, Department of Pharmaceutical Chemistry, Faculty of Pharmacy, Anadolu University, more, the concurrent NADPH deprivation impairs the 26470 Eskişehir, Turkey, e-mail: [email protected] activity of other NADPH-dependent and causes * Corresponding author: Şükrü Beydemir, Department of an imbalance between the generation of intracellular Biochemistry, Faculty of Pharmacy, Anadolu University, 26470 reactive oxygen species and cellular antioxidant defense. ş Ş Eski ehir, Turkey; The Rectorate of Bilecik eyh Edebali University, Concomitantly, pseudohypoxia, which results from the 11230 Bilecik, Turkey, e-mail: [email protected] + - Belgin Sever, Kaan Özbaş, Ahmet Özdemir: Department of NAD depletion during the oxidation of sorbitol to fruc Pharmaceutical Chemistry, Faculty of Pharmacy, Anadolu University, tose by , causes further metabolic 26470 Eskişehir, Turkey and signaling alterations by exacerbating redox imbal- Yeliz Demir: Department of Pharmacy Services, Nihat Delibalta Göle ance. Fructose, the end of the polyol pathway, Vocational High School, Ardahan University, 75700 Ardahan, Turkey is more reactive than glucose as a glycating agent; and Cüneyt Türkeş: Department of Biochemistry, Faculty of Pharmacy, Erzincan Binali Yıldırım University, 24100 Erzincan, Turkey the increased formation of fructose also gives rise to Gülşen Akalın Çiftçi: Department of Biochemistry, Faculty of pathological conditions by promoting protein glycation Pharmacy, Anadolu University, 26470 Eskişehir, Turkey and the formation of advanced glycation end products

Open Access. © 2021 Belgin Sever et al., published by De Gruyter. This work is licensed under the Creative Commons Attribution 4.0 International License. 348  Belgin Sever et al. and thus leading to alterations in protein functions. Apart were also carried out to estimate their physicochemical from its role in T2D complications, AR is an important parameters for the evaluation of their oral bioavailability mediator in oxidative and inflammatory-signaling path- and drug likeness. ways implicated in the pathophysiology of cardiovas- cular disorders, sepsis, and cancer. In this context, AR is identified as a multidisease target for the design of potent agents able to counteract the development of 2 Experimental section long-term T2D complications as well as nondiabetic dis- [ – ] eases 5 13 . 2.1 Chemistry The recent findings related to the pathophysiological role of AR have led to the discovery of a great variety of AR inhibitors so far, and most of them have been evalu- 2.1.1 General ated in preclinical and clinical trials. However, their development is mostly hampered by low in vivo potency, 2,4-TZD and urea were procured from Acros Organics adverse effects, or pharmacokinetic drawbacks [5–13]. (Geel, Belgium) and VWR Chemicals (Leuven, Belgium), 2,4-Thiazolidinedione (TZD) stands out as a privi- respectively. Aromatic aldehydes were purchased from leged scaffold for the identification of promising thera- Alfa Aesar (Karlsruhe, Germany) or Sigma-Aldrich (St. peutic agents for the management of T2D, targeting a Louis, MO, USA). Melting points (MPs) were detected plethora of crucial enzymes/receptors such as peroxi- using Electrothermal IA9200 MP apparatus (Staffordshire, some proliferator-activated receptor gamma, AR, protein UK). Infrared (IR), nuclear magnetic resonance (NMR; tyrosine phosphatase 1B, and so on [14–26]. In the search 1Hand13C), mass spectra, and elemental analyses for novel AR inhibitors, 2,4-TZDs are of great importance were recorded on IRPrestige-21 FT-IR spectrophotometer as safer bioisosteres of hydantoin, which is considered (Shimadzu, Tokyo, Japan), Varian 400 MHz FT-NMR spec- as the main cause of hypersensitivity reactions provoked trometer (Agilent, Palo Alto, CA, USA), VG Quattro Mass by some AR inhibitors, such as sorbinil. Currently, only spectrometer (Agilent, Minnesota, USA), and Perkin Elmer epalrestat (EPR) bearing a 2-thioxo-4-thiazolidinone EAL 240 elemental analyzer (Perkin-Elmer, Norwalk, CT, scaffold (Figure 1) is commercially available in few Asian USA),respectively. countries (such as Japan and India) as an AR inhibitor approved for the management of diabetic neuropathy. This agent is able to slow the progression of diabetic 2.1.2 Synthesis of 5-(arylidene)thiazolidine-2,4- neuropathy and ameliorate its symptoms without any diones (1–8) serious side effects after long-term use. However, further long-term comparative studies should be carried out to A mixture of aromatic aldehyde (2 mmol) and 2,4-TZD elucidate its efficacy in different patient populations [5]. (2 mmol) in the presence of urea (20 mmol) was heated Taking into account the knowledge obtained so far in an oil bath at 150°C for 2 h. Upon completion of the [5–26] and the potential of TZD-based small molecules as reaction, it was then dispersed with hot water and col- AR inhibitors [14–26], herein we reported the preparation lected by filtration. The product was crystallized from of new 2,4-TZDs and in vitro studies related to their AR ethanol [27]. inhibitory activities and cytotoxicity toward L929 mouse fibroblast cell line. In an effort to explore their possible binding modes in the binding site of AR, molecular docking studies were performed. In silico absorption, dis- 2.1.2.1 5-(2-Fluoro-4-methoxybenzylidene)thiazolidine- tribution, metabolism, and excretion (ADME) studies 2,4-dione (1)

Yield: 88%. MP: 215–220°C. −1 IR νmax (cm ): 3450.65, 3367.71, 3080.32, 3007.02, S S 2843.07, 1728.22, 1708.93, 1668.43, 1616.35, 1593.20, O N O 1573.91, 1506.41, 1436.97, 1388.75, 1342.46, 1313.52, OH 1294.24, 1269.16, 1251.80, 1190.08, 1161.15, 1093.64, 1028.06, 950.91, 871.82, 858.32, 825.53, 806.25, 756.10, Figure 1: Epalrestat. 729.09, 713.66, 692.44, and 646.15. 2,4-Thiazolidinediones as potent AR inhibitors  349

1 H NMR (400 MHz, DMSO-d6) δ (ppm): 3.80 (s, 3H), 2.1.2.4 5-(2-Hydroxy-5-methoxy-3-nitrobenzylidene) 6.38 (s, 1H), 6.86–6.89 (m, 1H), 7.68 (d, J = 9.6 Hz, 1H), thiazolidine-2,4-dione (4) ( ) ( ) 8.49 s, 1H , 10.41, and 11.18 2s,1H . – 13 Yield: 74%. MP: 245 246°C. C NMR (100 MHz, DMSO-d6) δ (ppm): 55.76 (CH ), 3 IR ν (cm−1): 3562.52, 3425.58, 3334.92, 3076.46, 101.61 (d, J = 25.0 Hz, CH), 110.79 (CH), 113.05 (d, J = max 2945.30, 1693.50, 1681.93, 1668.43, 1598.99, 1537.27, 1531.48, 12.8 Hz, C), 116.12 (C), 127.85 (CH), 130.23 (d, J = 3.8 Hz, 1454.33, 1344.38, 1307.74, 1242.16, 1197.79, 1166.93, 1095.57, CH), 155.41 (C), 160.85 (d, J = 11.5 Hz, C), 165.23 (C), and ( ) 1041.56, 929.69, 837.11, 752.24, and 665.44. 167.44 C . 1 H NMR (400 MHz, DMSO-d6) δ (ppm): 3.81 (s, 3H), MS (FAB) m/z 254.11 [M + H]+. 6.72–6.88 (m, 1H), 7.87–8.02 (m, 1H), 8.47 (s, 1H), 10.26 Anal. Calcd. for C H FNO S: C, 52.17; H, 3.18; and N, 11 8 3 (s, 1H), and 11.10 (brs, 1H). 5.53. Found: C, 52.15; H, 3.19; and N, 5.54. 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 55.78 (CH3), 108.20 (CH), 116.06 (C), 118.01 (CH), 118.42 (C), 135.55 (CH), 137.90 (C), 145.44 (C), 155.05 (C), 165.30 (C), and 167.40 (C). 2.1.2.2 5-(2-Fluoro-5-methoxybenzylidene)thiazolidine- MS (FAB) m/z 297.14 [M + H]+. 2,4-dione (2) Anal. Calcd. for C11H8N2O6S: C, 44.60; H, 2.72; and N, Yield: 55%. MP: 235–240°C. 9.46. Found: C, 44.63; H, 2.71; and N, 9.44. −1 IR νmax (cm ): 3466.08, 3045.60, 2945.30, 2835.36, 1728.22, 1666.50, 1593.20, 1496.76, 1463.97, 1419.61, 1382.96, 1323.17, 1282.66, 1211.30, 1099.43, 1031.92, 1012.63, 956.69, 2.1.2.5 5-(3-Chloro-4-fluorobenzylidene)thiazolidine- 877.61, 839.03, 804.32, 758.02, 711.73, 694.37, and 644.22. - ( ) 1 2,4 dione 5 H NMR (400 MHz, DMSO-d6) δ (ppm): 3.80 (s, 3H), 6.38 (s, 1H), 6.88–6.92 (m, 1H), 7.12–7.17 (m, 2H), 10.66, Yield: 68%. MP: 218–224°C. −1 and 11.18 (2 brs, 1H). IR νmax (cm ): 3444.87, 3387.00, 3037.89, 1737.86, 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 55.70 (CH3), 1708.93, 1668.43, 1591.27, 1504.48, 1435.04, 1346.31, 1263.37, 113.73 (d, J = 1.9 Hz, CH), 115.78 (CH), 115.88 (d, J = 3.2 Hz, 1195.87, 1126.43, 1091.71, 1060.85, 1024.20, 920.05, 898.83, CH), 116.12 (C), 121.23 (d, J = 14.1 Hz, C), 129.86 (d, J = 864.11, 819.75, 792.74, 756.10, 729.09, 709.80, 675.09, and 1.9 Hz, CH),155.53(C),160.79(C),165.13(C),and167.44(C). 642.30. + 1 MS (FAB) m/z 254.14 [M + H] . H NMR (400 MHz, DMSO-d6) δ (ppm): 6.37 (s, 1H),

Anal. Calcd. for C11H8FNO3S: C, 52.17; H, 3.18; and N, 7.39 (t, J = 8.8 Hz, 1H), 7.83 (d, J = 5.6 Hz, 1H), 8.49 (s, 1H), 5.53. Found: C, 52.19; H, 3.17; and N, 5.52. 10.67, and 11.25 (2 brs, 1H). 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 116.91 (C), 117.02 (d, J = 21.2 Hz, CH), 120.09 (d, J = 17.3 Hz, C), 128.70 (d, J = 1.9 Hz, CH), 130.15 (d, J = 7.0 Hz, CH), 2.1.2.3 5-(2-Hydroxy-3-methylbenzylidene)thiazolidine- 130.72 (C), 130.96 (d, J = 3.8 Hz, CH), 155.38 (C), 165.27 2,4-dione (3) (C), and 167.45 (C). Yield: 36%. MP: 168–170°C. MS (FAB) m/z 258.06 [M + H]+. −1 IR νmax (cm ): 3541.31, 3431.36, 3336.85, 3050.46, Anal. Calcd. for C10H5ClFNO2S: C, 61.78; H, 4.75; and 2924.09, 2858.51, 1693.50, 1681.93, 1600.92, 1525.69, N, 6.00. Found: C, 61.81; H, 4.73; and N, 6.01. 1469.76, 1435.04, 1381.03, 1265.30, 1188.15, 1095.57, 1039.63, 765.74, 746.45, 680.87, and 650.01. 1 H NMR (400 MHz, DMSO-d6) δ (ppm): 2.30 (s, 3H), 2.1.2.6 5-(3-Chloro-4-methylbenzylidene)thiazolidine- 6.66–7.44 (m, 3H), 8.50 (s, 1H), 10.12 (s, 1H), 10.41, and 2,4-dione (6) 11.18 (2s,1H). 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 15.23 (CH3), Yield: 83%. MP: 225–230°C. −1 116.10 (C), 116.40 (C), 122.43 (CH), 125.82 (CH), 126.11 (C), IR νmax (cm ): 3456.44, 3375.43, 3028.24, 2929.87, 129.86 (CH),131.20(CH),150.15(C),165.13(C),and167.44(C). 1728.22, 1710.86, 1660.71, 1589.34, 1496.76, 1440.83, 1392.61, MS (FAB) m/z 236.16 [M + H]+. 1361.74, 1334.74, 1284.59, 1253.73, 1219.01, 1188.15, 1124.50,

Anal. Calcd. for C11H9NO3S: C, 56.16; H, 3.86; and N, 1087.85, 1053.13, 1031.92, 997.20, 906.54, 887.26, 860.25, 5.95. Found: C, 56.13; H, 3.87; and N, 5.97. 821.68, 756.10, 711.73, 675.09, 644.22, and 632.65. 350  Belgin Sever et al.

1 H NMR (400 MHz, DMSO-d6) δ (ppm): 2.32 (s, 3H), 2.2 Biochemistry 6.35 (s, 1H), 7.33 (d, J = 8.0 Hz, 1H), 7.66 (s, 1H), 8.48 (s, 1H), 10.61, and 11.21 (2 brs, 1H). 2.2.1 AR activity assay 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 19.35 (CH3), 116.91 (C), 127.94 (CH), 128.29 (CH), 128.97 (CH), 131.29 Purification of sheep liver AR was done according to the (C), 132.52 (C), 133.80 (C), 135.55 (CH), 165.35 (C), and previous studies [28–32]. Bradford method was utilized to 167.45 (C). determine the quantitative protein [33]. The enzyme purity MS (FAB) m/z 254.17 [M + H]+. was checked according to Laemmli’sprocedure[34,35].AR

Anal. Calcd. for C11H8ClNO2S: C, 52.08; H, 3.18; and N, activity was spectrophotometrically evaluated based on the 5.52. Found: C, 52.05; H, 3.20; and N, 5.52. decrease in absorbance of NADPH at 340 nm [36–38].

2.2.2 In vitro inhibition studies 2.1.2.7 5-(3-Methoxy-2-nitrobenzylidene)thiazolidine- 2,4-dione (7) The AR activity was determined in the presence of different

concentrations of compounds 1–8.TheIC50 value of each Yield: 43%. MP: 190–195°C. compound was calculated from activity%–[Compound] −1 IR νmax (cm ): 3323.35, 3022.45, 2941.44, 2837.29, graphs [36]. For determining the inhibition types of the com- 1658.78, 1651.07, 1598.99, 1537.27, 1454.33, 1402.25, 1359.82, pounds, Lineweaver–Burk graph was plotted [39] according 1261.45, 1178.51, 1118.71, 1068.56, 1039.63, 995.27, 852.54, to the previous works [40–42]. 732.95, and 642.30. 1 H NMR (400 MHz, DMSO-d6) δ (ppm): 3.83 (s, 3H), 6.85–7.90 (m, 3H), 8.69 (s, 1H), and 11.10 (brs, 1H). 2.2.3 Cell culture and drug treatment 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 55.76 (CH3), 114.40 (CH), 116.50 (C), 119.69 (CH), 128.90 (C), 131.30 (C), The L929 mouse fibroblast cells (ATCC, CCL-1TM)(Manassas, 134.99 (CH), 136.56 (CH), 155.41 (C), 165.36 (C), and VA, USA) were cultured and drug treatments were per- 167.47 (C). formed as previously reported [42]. MS (FAB) m/z 281.16 [M + H]+.

Anal. Calcd. for C11H8N2O5S: C, 47.14; H, 2.88; and N, 10.00. Found: C, 47.11; H, 2.89; and N, 10.02. 2.2.4 MTT assay

MTT test was performed to examine the cytotoxic effects of compounds 1–8 on the L929 cells as previously explained 2.1.2.8 5-(5-Chloro-2-hydroxy-3-methylbenzylidene) [43] but with minor modifications [42]. thiazolidine-2,4-dione (8) The percentage of the viable cells was calculated using the following formula: (%)=[100 ×(sample absor- Yield: 40%. MP: 160–165°C. −1 bance)/(control absorbance)]. IR νmax (cm ): 3541.31, 3454.51, 3334.92, 3076.46, 2927.94, 1693.50, 1681.93, 1598.99, 1566.20, 1514.12, 1469.76, 1435.04, 1381.03, 1311.59, 1192.01, 1120.64, 1043.49, 900.76, 864.11, 752.24, 731.02, 651.94, and 634.58. 2.2.5 Statistical studies 1 H NMR (400 MHz, DMSO-d6) δ (ppm): 2.26 (s, 3H), 6.92–7.68 (m, 2H), 8.51 (s, 1H), 10.21 (s, 1H), and 11.10 GraphPad Prism version 7 for Mac (GraphPad Software, (brs, 1H). La Jolla, CA, USA) was used for data analysis and graphs. 13 C NMR (100 MHz, DMSO-d6) δ (ppm): 15.25 (CH3), SigmaPlot version 12 for Windows (Systat Software, 116.61 (C), 117.89 (C), 124.78 (CH), 126.76 (C), 129.39 (CH), San Jose, CA, USA) was performed to calculate the inhi- 132.80 (C),135.56(CH),150.15(C),165.31(C),and167.42(C). bition constants. The fit of enzyme inhibition models was MS (FAB) m/z 270.12 [M + H]+. compared using the extra sum-of-squares F test and the

Anal. Calcd. for C11H8ClNO3S: C, 48.99; H, 2.99; and Akaike’s corrected Information Criterion (AICc) approach. N, 5.19. Found: C, 48.98; H, 2.97; and N, 5.21. The results were expressed as mean ± standard error of the 2,4-Thiazolidinediones as potent AR inhibitors  351

O mean (95% confidence intervals).Differences between H O CHO data sets were considered statistically significant when N i R R NH the p value was less than 0.05. O S S O

Scheme 1: The synthetic route for the preparation of compounds – ( ) 2.3 Molecular docking studies 1 8. Reagents and conditions: i NH2CONH2, oil bath, 150°C, 2 h.

Molecular docking simulations were conducted using panels analyses. In their IR spectra, the C]O stretching vibra- (LigPrep [44],Maestro[45], Prime molecular mechanics– tions resulted in the formation of two characteristic generalized Bornsurfacearea[MM-GBSA][46],ProteinPre- bands at 1737.86–1651.07 cm−1.TheN–Hstretchingvibra- paration Wizard [47], and Receptor Grid Generation [48]) in tion belonging to the N–H proton of the thiazolidine scaffold theSchrödingerSuite2020-2forMac.Thehigh-resolution gave rise to the bands in the region 3466.08–3323.35 cm−1.In 3D crystal structure of AR (PDB ID: 4JIR; 2.00 Å)[49] was the IR spectra of compounds 3, 4,and8,theO–Hstretch- downloaded from RCSB Protein Data Bank [50] andusedfor ing band appeared at 3562.52–3541.31 cm−1.Inthe1H molecular docking. Protein Preparation Wizard [51] was NMR spectra of all compounds except compound 2,the used to prepare the crystal structure [52].The3Dstructures signal due to the benzylidene CH proton was observed at of compounds 1–8 were sketched using ChemDraw Pro ver- 8.47–8.69 ppm as a singlet. In the 1HNMRspectraofcom- sion 19.1 for Mac [53](PerkinElmer, Inc., Waltham, MA, pounds 4, 7,and8,thesignalduetotheN–Hproton USA). The molecules were subjected to ligand preparation appeared at 11.10 ppm as a broad singlet, whereas in the using LigPrep tool [54] in default conditions at pH 7.4 ± 0.5 1H NMR spectra of other compounds, the signal due to the [55] with Epik [56] intheOPLS3eforcefield [57,58]. Receptor N–H proton appeared in the region 10.41–11.25 ppm as two Grid Generation tool [59] was used to generate the grid for singlets or broad singlets. In the 1H NMR spectra of methoxy- docking. Binding sites were defined using cocrystallized nat- substituted compounds 1, 2, 4,and7, the signal due to the ural ligand (EPR).Glideextraprecision(Glide XP)[60,61] methoxy protons was observed in the region 3.80–3.83 ppm was used for docking [62]. Docked poses were rescored using as a singlet. In the 1H NMR spectra of methyl-substituted the MM-GBSA approach [63,64]. compounds 3, 6,and8, the signal due to the methyl protons occurred in the region 2.26–2.32 ppm as a singlet. The O–H proton gave rise to a singlet peak at 10.12–10.26 ppm in 1 3 4 8 13 In silico the H NMR spectra of compounds , ,and .Inthe C 2.4 ADME studies NMR spectra of compounds 1–8, the signals due to the car- bons of two C]Ogroupswereobservedintheregion QikProp, a predictive ADME module within the Maestro 165.13–167.47 ppm. The signal due to the benzylidene suite produced by Schrödinger, was performed to predict CH carbon appeared at 129.86–136.56 ppm. In the 13C 1–8 the ADME properties of compounds . NMR spectra of methoxy-substituted compounds 1, 2, 4,and7, the methoxy carbon gave rise to the peak at Ethical approval: The conducted research is not related to 55.70–55.78 ppm. In the 13C NMR spectra of methyl-sub- either human or animal use. stituted compounds 3, 6,and8, the signal due to the methyl carbon occurred in the region 15.23–19.35 ppm.

3 Results and discussion 3.2 In vitro AR inhibitory activity and cytotoxicity 3.1 Chemistry

The IC50, Ki, and inhibition types of compounds 1–8 The preparation of the hitherto unreported 5-(arylidene) were determined to investigate their ability to inhibit thiazolidine-2,4-diones (1–8) was carried out through the AR (Table 1). According to in vitro data, compounds 1–8 solvent-free reaction of 2,4-TZD with aromatic aldehydes showed inhibitory effects on AR, with IC50 values ranging in the presence of urea (Scheme 1). from 0.273 to 0.533 µM and Ki values ranging from 0.445 The structures of compounds 1–8 were verified by to 0.943 µM (Figure 2). Compounds 1, 3, and 5 were found IR, 1H NMR, 13C NMR, mass spectrometry and elemental to act as uncompetitive AR inhibitors, whereas other 352  Belgin Sever et al.

Table 1: AR inhibition data of compounds 1–8

O

R NH S O

2 2 Compound R IC50 (µM) R Ki (µM) R Inhibition type

1 2-F-4-OCH3 0.375 ± 0.007 0.9970 0.462 ± 0.015 0.9964 Uncompetitive

2 2-F-5-OCH3 0.273 ± 0.004 0.9981 0.549 ± 0.023 0.9923 Non-competitive

3 2-OH-3-CH3 0.382 ± 0.010 0.9955 0.445 ± 0.013 0.9982 Uncompetitive

4 2-OH-5-OCH3-3-NO2 0.349 ± 0.003 0.9994 0.769 ± 0.017 0.9977 Non-competitive 5 3-Cl-4-F 0.398 ± 0.011 0.9920 0.892 ± 0.020 0.9983 Uncompetitive

6 3-Cl-4-CH3 0.411 ± 0.005 0.9991 0.729 ± 0.019 0.9957 Non-competitive

7 3-OCH3-2-NO2 0.455 ± 0.008 0.9963 0.713 ± 0.014 0.9976 Non-competitive

8 2-OH-3-CH3-5-Cl 0.533 ± 0.015 0.9935 0.943 ± 0.024 0.9960 Non-competitive

The test results were indicated as mean ± standard deviation.

(a) (b)

Figure 2: (a) Percentage activity versus inhibitor concentration graph of compound 3 at five different concentrations. (b) Lineweaver–Burk plot of compound 3. compounds were identified as noncompetitive AR inhibi- therapeutic agents endowed with favorable pharmaco- tors. The order of 2,4-TZD-based AR inhibitors (1–8) kinetic profiles as well as devoid of severe unwanted effects (from the most active to the least active) according to is an uphill task for researchers [5]. On this basis, herein their Ki values was noted to be as follows: compound 3 > compound 1 > compound 2 > compound 7 > compound 6 > 4 > 5 > 8 compound compound compound . The Table 2: IC50 values of compounds 1–8 for L929 cell line after 24 h in vitro data pointed out the significance of the arylidene incubation group at the fifth position of 2,4-TZD scaffold. The intro- ( )a duction of a chlorine group into the fifth position of the Compound IC50 µM benzylidene moiety of compound 3 (IC50 = 0.382 ± 0.010 µM, 1 0.55 ± 0.07 Ki = 0.445 ± 0.013 µM) led to a substantial decrease 2 2.63 ± 0.15 3 8.9 ± 0.66 in AR inhibitory potency (IC50 = 0.533 ± 0.015 µM, 4 2.25 ± 0.35 Ki = 0.943 ± 0.024 µM for compound 8) and an alteration 5 3.47 ± 0.84 in inhibition type (the inhibition type of compound 3 was 6 1.43 ± 0.06 8 uncompetitive, while the inhibition type of compound 7 >25 was noncompetitive). 8 >25 In the search for AR inhibitors for the management of T2D and its complications, the identification of potent a Results were given as mean ± SD. 2,4-Thiazolidinediones as potent AR inhibitors  353

the cytotoxic activities of compounds 1–8 against L929 value of compound 1 for L929 cells (IC50 = 0.55 ± 0.07 µM) mouse fibroblast (healthy) cells were determined (Table 2). was slightly close to its IC50 value for AR inhibition

According to the MTT assay, all compounds did not show (IC50 = 0.375 ± 0.007 µM). The effects of compound 1 on significant cytotoxicity toward L929 cells at their effective percentages of L929 cell viability at different concentra- concentrations. The IC50 values of compounds 2–6 for tions (0.39, 0.78, and 6.26 µM) were found as 68.85 ± 2.47, L929 cells were found to be between 1.43 and 8.9 µM, 38.61 ± 14.71, and 31.60 ± 3.15, respectively (p < 0.05). while the IC50 values of compounds 7 and 8 for L929 These results showed that compound 1 caused dose- cell line were higher than 25 µM, pointing out the safety dependent cytotoxicity on L929 cell line at tested concen- of compounds 2–8 as AR inhibitors. However, the IC50 trations. On the other hand, the effects of compound 3

Figure 3: Interactions of the ligands with the key amino acids within the binding site of AR (PDB ID: 4JIR, 2.00 Å). (a) 2D ligand interaction diagram of 4JIR with native ligand EPR (epalrestat). (b) 2D ligand interaction diagram of 4JIR with compound 3. 354  Belgin Sever et al. on percentages of L929 cell viability at different concen- According to Zhang et al. [49], the native ligand EPR trations (0.78, 6.26, and 12.5 µM) were determined as (with an MM-GBSA value of –33.42 kcal/mol and a docking 92.22 ± 5.43, 91.99 ± 13.69, and 22.72 ± 2.39, respectively score of –7.19) formed three H-bonds with Tyr48, His110, (p < 0.05). The cytotoxicity of compound 3 was low at and Trp111 (with distances 1.56, 1.83, and 2.25 Å, respec- 0.78 and 6.26 µM. This outcome indicated that compound tively), in the catalytic domain of 4JIR. Moreover, the 3 displayed low cytotoxicity toward L929 cell line at its most prominent amino acid residues accommodating

IC50 value for AR inhibition (IC50 = 0.382 ± 0.010 µM). hydrophobic fragments were Tyr48 and Trp111 as well as Trp20, Val47, Trp79, Phe122, Trp219, and Cys298. Com- pound 3 (with an MM-GBSA value of –33.69 kcal/mol and a docking score of –7.19) exhibited two interactions with 3.3 Molecular docking studies the target enzyme. Asn160 and Cys298 showed H-acceptor interactions with the hydroxyl and carbonyl groups (with In an effort to gain insight into the binding mode of the distances 2.10 and 2.41 Å, respectively) of the ligand. The TZD scaffold in the enzyme binding site, compound 3 weak π–π stacking between Trp20, His110, and Tyr209 carrying 2-hydroxy-3-methylbenzylidene moiety at the residues and the benzylidene moiety was observed in fifth position of 2,4-TZD scaffold was docked into the the 2D ligand interaction diagram (Figure 3).Theresults binding site of AR (PDB ID: 4JIR) as a representative of provided insights into the interactions between compounds compounds 1–8 (Figure 3). The best poses for ligand 1–8 and AR and rationalized the experimental data. according to score and binding interactions were refined by the MM-GBSA-based approach to assess the electro- static contribution of the variable dielectric surface gene- ralized born (VSGB) solvation model. For validation of 3.4 In silico pharmacokinetic studies molecular docking simulations, the cocrystalized ligand EPR was extracted and redocked into the binding site. The discovery of AR inhibitors with favorable ADME profiles Their root mean square deviation (RMSD) score was com- represents a crucial turning point in the challenging route puted to evaluate the quality of the cocrystallized ligand. to develop a new generation of safer AR inhibitors [5].On The results were compared with EPR derived from the this basis, the pharmacokinetic features of compounds 1–8 corresponding 4JIR, and the docking poses were super- were assessed by means of QikProp (Table 3). According imposed. The RMSD score was found as 1.20 Å. The docking to in silico prediction, their SASA, QPlogPo/w, CIQPlogS, pattern of EPR was compared with that of compound 3. and QPlogKhsa values were within the recommended

Table 3: Predicted ADME profiles of compounds 1–8

Compound SASAa QPlogPo/wa CIQPlogSa QPlogKhsaa Human oral absorption%b Rule of fivec Rule of threec

1 428.912 1.786 −3.528 −0.236 86.866 0 0 2 421.221 1.746 −3.528 −0.240 86.630 0 0 3 423.671 1.156 −3.130 −0.293 77.063 0 0 4 465.209 0.375 −3.705 −0.406 56.281 0 0 5 417.076 2.184 −3.892 −0.145 89.199 0 0 6 436.250 2.144 −3.813 −0.044 88.965 0 0 7 456.458 1.058 −3.691 −0.304 70.607 0 0 8 448.637 1.694 −3.791 −0.188 81.474 0 0 a Recommended values for the total solvent accessible surface area (SASA): 300–1,000 Å2; the predicted octanol/water partition coefficient (QPlogPo/w): −2 to 6.5; the conformation-independent predicted aqueous solubility (CIQPlogS): −6.5 to 0.5; the predicted binding to human serum albumin (QPlogKhsa): −1.5 to 1.5.b Predicted human oral absorption on 0 to 100% scale. Human oral absorption higher than 80% is considered to be high, while human oral absorption lower than 25% is considered to be poor.c Rule of Five: Number of violations of Lipinski’s rule of five. The rules are molecular weight of the molecule < 500, QPlogPo/w < 5, hydrogen-bond donor atoms ≤5, hydrogen-bond acceptor atoms ≤ 10. Compounds that provide these rules are considered as drug-like molecules. Rule of Three: Number of violations of Jorgensen’s rule of three. The three rules are QPlogS (predicted aqueous solubility)>−5.7, QPPCaco > 22 nm/s, # primary metabolites < 7. Compounds with fewer (and preferably no) violations of these rules are more likely to be orally available agents (Schrödinger Release 2016- 2: Schrödinger, LLC, New York, NY, USA). 2,4-Thiazolidinediones as potent AR inhibitors  355 range. Their human oral absorption percentages were methodology, project administration, resources, software, found to range from 56.281 to 89.199%. All compounds validation, visualization, writing of the original draft, comply with Lipinski’sruleoffive and Jorgensen’srule reviewing, and editing; KÖ contributed to investigation, of three, and therefore they are expected to have favor- methodology, writing, reviewing, and editing; GAÇ was able oral bioavailability and drug-like features. in charge of data curation, formal analysis, methodology, software, validation, writing, reviewing, and editing; ŞB was involved in conceptualization, methodology, software, project administration, resources, writing, reviewing, and 4 Conclusions editing; AÖ contributed to conceptualization, methodology, software, resources, writing, reviewing, and editing. In conclusion, a facile and versatile synthetic procedure was performed to obtain new 5-(arylidene)thiazolidine- Conflict of interest: Belgin Sever, the coauthor of this 2,4-diones, which were evaluated for their AR inhibitory article, is a current Editorial Board member of Open effects and cytotoxic effects on L929 cells. According to Chemistry. This fact did not affect the peer-review pro- in vitro assay, compounds 1–8 showed inhibitory effects cess. The authors declare no other conflict of interest. on AR with Ki values ranging from 0.445 to 0.943 µM. - Taking into account the Ki values, compound 3 was found Data availability statement: All data generated or ana as the most promising AR inhibitor with a Ki value of lyzed during this study are included in this published 0.445 ± 0.013 µM and its inhibition type was determined article [and its supplementary information files]. as uncompetitive. The MTT assay indicated that com- pound 3 did not exert cytotoxicity toward L929 cells at its effective concentration. Based on molecular docking studies, compound 3 interacted with crucial amino acid References residues in the binding site of AR (PDB ID: 4JIR). Taking into account in silico ADME studies, this compound is [1] Schäfer SA, Machicao F, Fritsche A, Häring H-U, Kantartzis K. expected to have a favorable pharmacokinetic profile. New type 2 diabetes risk genes provide new insights in insulin In vitro in silico and studies pointed out the potential of secretion mechanisms. Diabetes Res Clin Pract. 2011;93:S9–24. compound 3 as an orally bioavailable AR inhibitor for the [2] Liu Y, Hu Y, Liu T. Recent advances in non-peptidomimetic management of T2D complications as well as nondiabetic dipeptidyl peptidase 4 inhibitors: medicinal chemistry and diseases. preclinical aspects. Curr Med Chem. 2012;19:3982–99. [3] Kerru N, Singh-Pillay A, Awolade P, Singh P. Current anti-dia- betic agents and their molecular targets: a review. Eur J Med Funding: This work was supported by the Research Fund Chem. 2018;152:436–88. of Anadolu University (grant numbers: 2005S019 and [4] De S, Banerjee S, Kumar SKA, Paira P. Critical role of dipeptidyl 1610S681), the Research Fund of Ardahan University peptidase IV: a therapeutic target for diabetes and cancer. (grant number: 2019-008), and the Research Fund of Mini Rev Med Chem. 2019;19:88–97. [ ] Erzincan Binali Yıldırım University (grant number: FBA- 5 Maccari R, Ottanà R. Targeting aldose reductase for the treatment of diabetes complications and ınflammatory dis- 2017-501). eases: new insights and future directions. J Med Chem. 2015;58:2047–67. Author contributions: BS was involved in conceptualiza- [6] Costantino L, Rastelli G, Gamberini MC, Barlocco D. tion, investigation, formal analysis, methodology, soft- Pharmacological approaches to the treatment of diabetic ware, writing of the original draft, reviewing, and editing; complications. Exp Opin Ther Pat. 2000;10:1245–62. [7] Brownlee M. Biochemistry and molecular cell biology of MDA was in charge of conceptualization, data curation, diabetic complications. Nature. 2001;414:813–20. - formal analysis, funding acquisition, investigation, meth [8] Lorenzi M. The polyol pathway as a mechanism for diabetic odology, project administration, resources, software, retinopathy: attractive, elusive, and resilient. Exp Diabetes writing – writing of the original draft, reviewing, and Res. 2007;2007:61038. editing; YD contributed to data curation, formal analysis, [9] Ramana KV. Aldose reductase: new insights for an old enzyme. – funding acquisition, methodology, project administration, BioMol Concepts. 2011;2:103 14. [10] Grewal AS, Bhardwaj S, Pandita D, Lather V, Sekhon BS. resources, software, validation, visualization, writing of Updates on aldose reductase inhibitors for management of the original draft, reviewing, and editing; CT was involved diabetic complications and non-diabetic diseases. Mini Rev in data curation, formal analysis, funding acquisition, Med Chem. 2016;16:120–62. 356  Belgin Sever et al.

[11] Sangshetti JN, Chouthe RS, Sakle NS, Gonjari I, Shinde DB. [27] Shah S, Singh B. Urea/thiourea catalyzed, solvent-free Aldose reductase: a multi-disease target. Curr Enzym Inhib. synthesis of 5-arylidenethiazolidine-2,4-diones and 5-aryli- 2014;10:2–12. dene-2-thioxothiazolidin-4-ones. Bioorg Med Chem Lett. [12] Quattrini L, La Motta C. Aldose reductase inhibitors: 2013- 2012;22:5388–91. present. Expert Opin Ther Pat. 2019;29:199–213. [28] Aslan HE, Beydemir Ş. Phenolic compounds: the inhibition [13] Dowarah J, Singh VP. Anti-diabetic drugs recent approaches effect on polyol pathway enzymes. Chem Biol Interact. and advancements. Bioorg Med Chem. 2020;28:115263. 2017;266:47. [14] Jain AK, Vaidya A, Ravichandran V, Kashaw SK, Agrawal RK. [29] Demir Y, Duran HE, Durmaz L, Taslimi P, Beydemir Ş, Gulçin İ. Recent developments and biological activities of thiazolidi- The influence of some nonsteroidal anti-inflammatory drugs none derivatives: a review. Bioorg Med Chem. on metabolic enzymes of aldose reductase, sorbitol dehydro- 2012;20:3378–95. genase, and α-glycosidase: a perspective for metabolic dis- [15] Jain VS, Vora DK, Ramaa CS. Thiazolidine-2,4-diones: Progress orders. Appl Biochem Biotechnol. 2020;190:437–47. towards multifarious applications. Bioorg Med Chem. [30] Demir Y, Işık M, Gülçin İ, Beydemir Ş. Phenolic compounds 2013;21:1599–620. inhibit the aldose reductase enzyme from the sheep kidney. [16] Chadha N, Bahia MS, Kaur M, Silakari O. Thiazolidine-2,4- J Biochem Mol Toxicol. 2017;31:e21936. dione derivatives: programmed chemical weapons for key [31] Şengül B, Beydemir Ş. The interactions of cephalosporins on protein targets of various pathological conditions. Bioorg Med polyol pathway enzymes from sheep kidney. Arch Physiol Chem. 2015;23:2953–74. Biochem. 2018;124:35–44. [17] Naim MJ, Alam MJ, Ahmad S, Nawaz F, Shrivastava N, Sahu M, [32] Demir Y, Durmaz L, Taslimi P, Gulçin İ. Antidiabetic properties et al. Therapeutic journey of 2,4-thiazolidinediones as a ver- of dietary phenolic compounds: inhibition effects on α-amy- satile scaffold: an insight into structure activity relationship. lase, aldose reductase, and α-glycosidase. Biotechnol Appl Eur J Med Chem. 2017;129:218–50. Biochem. 2019;66:781–6. [18] Kaur Manjal S, Kaur R, Bhatia R, Kumar K, Singh V, Shankar R, [33] Bradford MM. A rapid and sensitive method for the quantita- et al. Synthetic and medicinal perspective of thiazolidinones: tion of microgram quantities of protein utilizing the principle a review. Bioorg Chem. 2017;75:406–23. of protein-dye binding. Anal Biochem. 1976;72:248–54. [19] Verma SK, Yadav YS, Thareja S. 2,4-Thiazolidinediones as PTP [34] Laemmli UK. Cleavage of structural proteins during the 1B inhibitors: a mini review (2012–2018). Mini Rev Med Chem. assembly of the head of bacteriophage T4. Nature. 2019;19:591–8. 1970;227:680–5. [20] Bansal G, Thanikachalam PV, Maurya RK, Chawla P, [35] Demir Y, Köksal Z. Some sulfonamides as aldose reductase Ramamurthy S. An overview on medicinal perspective of thia- inhibitors: Therapeutic approach in diabetes. Arch Physiol zolidine-2,4-dione: a remarkable scaffold in the treatment of Biochem. 2020. doi: 10.1080/13813455.2020.1742166. in type 2 diabetes. J Adv Res. 2020;23:163–205. press. [21] Bruno G, Costantino L, Curinga C, Maccari R, Monforte F, [36] Cerelli MJ, Curtis DL, Dunn JP, Nelson PH, Peak TM, Nicolò F, et al. Synthesis and aldose reductase inhibitory Waterbury LD. Antiinflammatory and aldose reductase inhibi- activity of 5-arylidene-2,4-thiazolidinediones. Bioorg Med tory activity of some tricyclic arylacetic acids. J Med Chem. Chem. 2002;10:1077–84. 1986;29:2347–51. [22] Maccari R, Del Corso A, Giglio M, Moschini R, Mura U, Ottanà R. [37] Türkeş C, Demir Y, Beydemir Ş. Anti-diabetic properties of In vitro evaluation of 5-arylidene-2-thioxo-4-thiazolidinones calcium channel blockers: inhibition effects on aldose reduc- active as aldose reductase inhibitors. Bioorg Med Chem Lett. tase enzyme activity. Appl Biochem Biotechnol. 2011;21:200–3. 2019;189:318–29. [23] Maccari R, Vitale RM, Ottanà R, Rocchiccioli M, Marrazzo A, [38] Sever B, Altıntop MD, Demir Y, Akalın Çiftçi G, Beydemir Ş, Cardile V, et al. Structure–activity relationships and molecular Özdemir A. Design, synthesis, in vitro and in silico investiga- modelling of new 5-arylidene-4-thiazolidinone derivatives as tion of aldose reductase inhibitory effects of new thiazole- aldose reductase inhibitors and potential anti-inflammatory based compounds. Bioorg Chem. 2020;102:104110. agents. Eur J Med Chem. 2014;81:1–14. [39] Lineweaver H, Burk D. The determination of enzyme dissocia- [24] Chadha N, Silakari O. Identification of low micromolar dual tion constants. J Am Chem Soc. 1934;56:658–66. inhibitors for aldose reductase (ALR2) and poly (ADP-ribose) [40] Türkeş C, Söyüt H, Beydemir Ş.Effect of calcium channel polymerase (PARP-1) using structure based design approach. blockers on paraoxonase-1 (PON1) activity and oxidative Bioorg Med Chem Lett. 2017;27:2324. stress. Pharmacol Rep. 2014;66:74–80. [25] Maccari R, Del Corso A, Paoli P, Adornato I, Lori G, Balestri F, [41] Demir Y. The behaviour of some antihypertension drugs on et al. An investigation on 4-thiazolidinone derivatives as dual human serum paraoxonase-1: an important protector enzyme inhibitors of aldose reductase and protein tyrosine phospha- against atherosclerosis. J Pharm Pharmacol. 2019;71:1576–83. tase 1B, in the search for potential agents for the treatment of [42] Sever B, Altıntop MD, Demir Y, Pekdoğan M, Akalın Çiftçi G, type 2 diabetes mellitus and its complications. Bioorg Med Beydemir Ş, et al. An extensive research on aldose reductase Chem Lett. 2018;28:3712–20. inhibitory effects of new 4H-1,2,4-triazole derivatives. J Mol [26] Kousaxidis A, Petrou A, Lavrentaki V, Fesatidou M, Nicolaou I, Struct. 2021;1224:129446. Geronikaki A. Aldose reductase and protein tyrosine phos- [43] Mosmann T. Rapid colorimetric assay for cellular growth and phatase 1B inhibitors as a promising therapeutic approach for survival: application to proliferation and cytotoxicity assays. diabetes mellitus. Eur J Med Chem. 2020;207:112742. J Immunol Methods. 1983;65:55–63. 2,4-Thiazolidinediones as potent AR inhibitors  357

[44] Türkeş C. Investigation of potential paraoxonase-I inhibitors silico studies of novel 1,3-diaryltriazene-substituted sul- by kinetic and molecular docking studies: chemotherapeutic fathiazole derivatives. Arch Pharm. 2020;353:e2000102. drugs. Protein Pept Lett. 2019;26:392–402. [56] Taslimi P, Işık M, Türkan F, Durgun M, Türkeş C, Gülçin İ, et al. [45] Demir Y, Türkeş C, Beydemir Ş. Molecular docking studies and Benzenesulfonamide derivatives as potent acetylcholines- inhibition properties of some antineoplastic agents against terase, α-glycosidase, and glutathione S- inhibi- paraoxonase-I. Anti-cancer Agents Med Chem. tors: biological evaluation and molecular docking studies. 2020;20:887–96. J Biomol Struct Dyn. 2020. doi: 10.1080/ [46] Işık M, Beydemir Ş, Demir Y, Durgun M, Türkeş C, Nasır A, et al. 07391102.2020.1790422. in press. Benzenesulfonamide derivatives containing imine and amine [57] Türkeş C, Demir Y, Beydemir Ş. Some calcium-channel groups: inhibition on human paraoxonase and molecular blockers: kinetic and in silico studies on paraoxonase-I. docking studies. Int J Biol Macromol. 2020;146:1111–23. J Biomol Struct Dyn. 2020. doi: 10.1080/ [47] Beydemir Ş, Türkeş C, Yalçın A. Gadolinium-based contrast 07391102.2020.1806927. in press. agents: In vitro paraoxonase 1 inhibition, in silico studies. [58] Gündoğdu S, Türkeş C, Arslan M, Demir Y, Beydemir Ş. New Drug Chem Toxicol. 2019. doi: 10.1080/ isoindole‐1,3‐dione substituted sulfonamides as potent inhi- 01480545.2019.1620266. in press. bitors of carbonic anhydrase and acetylcholinesterase: [48] Türkeş C. A potential risk factor for paraoxonase 1: in silico and Design, synthesis, and biological evaluation. ChemistrySelect. in vitro analysis of the biological activity of proton-pump 2019;4:13347–55. inhibitors. J Pharm Pharmacol. 2019;71:1553–64. [59] Türkeş C, Demir Y, Beydemir Ş. Calcium channel blockers: [49] Zhang L, Zhang H, Zhao Y, Li Z, Chen S, Zhai J, et al. Inhibitor Molecular docking and inhibition studies on carbonic anhy- selectivity between aldo–keto reductase superfamily members drase I and II isoenzymes. J Biomol Struct Dyn. 2020. doi: AKR1B10 and AKR1B1: Role of Trp112 (Trp111). FEBS Lett. 10.1080/07391102.2020.1736631. in press. 2013;587:3681–6. [60] Lolak N, Akocak S, Türkeş C, Taslimi P, Işık M, Beydemir Ş, [50] Işık M, Demir Y, Durgun M, Türkeş C, Necip A, Beydemir Ş. et al. Synthesis, characterization, inhibition effects, and Molecular docking and investigation of 4-(benzylideneamino)- molecular docking studies as acetylcholinesterase, α-glyco- and 4-(benzylamino)-benzenesulfonamide derivatives as sidase, and carbonic anhydrase inhibitors of novel benzene- potent AChE inhibitors. Chem Pap. 2020;74:1395–405. sulfonamides incorporating 1,3,5-triazine structural motifs. [51] Türkeş C, Beydemir Ş.Küfrevioğlu Öİ. In vitro and in silico stu- Bioorg Chem. 2020;100:103897. dies on the toxic effects of antibacterial drugs as human serum [61] Kılıç A, Beyazsakal L, Işık M, Türkeş C, Necip A, Takım K, et al. paraoxonase 1 inhibitor. ChemistrySelect. 2019;4:9731–6. Mannich reaction derived novel boron complexes with amine- [52] Türkeş C. Inhibition effects of phenolic compounds on human bis(phenolate) ligands: synthesis, spectroscopy and in vitro/ serum paraoxonase-1 enzyme. J Inst Sci Technol. 2019;9:1013–22. in silico biological studies. J Organomet Chem. 2020;927:121542. [53] Durgun M, Türkeş C, Işık M, Demir Y, Saklı A, Kuru A, et al. [62] Türkeş C, Beydemir Ş. Inhibition of human serum paraoxo- characterisation, biological evaluation and in silico studies of nase-I with antimycotic drugs: in vitro and in silico studies. sulphonamide Schiff bases. J Enzyme Inhib Med Chem. Appl Biochem Biotechnol. 2020;190:252–69. 2020;35:950–62. [63] Istrefi Q, Türkeş C, Arslan M, Demir Y, Nixha AR, Beydemir Ş, [54] Türkeş C, Arslan M, Demir Y, Cocaj L, Nixha AR, Beydemir Ş. et al. Sulfonamides incorporating ketene N,S-acetal bioisos- Synthesis, biological evaluation and in silico studies of novel teres as potent carbonic anhydrase and acetylcholinesterase N-substituted phthalazine sulfonamide compounds as potent inhibitors. Arch Pharm. 2020;353:e1900383. carbonic anhydrase and acetylcholinesterase inhibitors. [64] Sever B, Türkeş C, Altıntop MD, Demir Y, Beydemir Ş. Thiazolyl- Bioorg Chem. 2019;89:103004. pyrazoline derivatives: in vitro and in silico evaluation as [55] Işık M, Akocak S, Lolak N, Taslimi P, Türkeş C, Gülçin İ, et al. potential acetylcholinesterase and carbonic anhydrase inhi- Synthesis, characterization, biological evaluation, and in bitors. Int J Biol Macromol. 2020;163:1970–88.